Unconditional Cash Transfer and Household Resilience: Results from the Malawi Cash Transfer Program

Transcription

1 Unconditional Cash Transfer and Household Resilience: Results from the Malawi Cash Transfer Program Sudhanshu Handa University of North Carolina at Chapel Hill shanda@ .unc.edu Frank Otchere* University of North Carolina at Chapel Hill fkotcher@ad.unc.edu * Author for correspondence

2 ABSTRACT The concept of resilience is increasingly gaining traction in the international development literature as a way to profile, rank and predict the response capacity of households to shocks and stressors to livelihoods, particularly those that threaten food security. The objective is to provide a more rigorous framework and a single reference indicator for the design and implementation of sustainable long-term development initiatives that minimize the need for perennial mobilization for humanitarian and emergency assistance. While there are still debates about the construct and measurement, there is general consensus that a household s resilience encompasses aspects of household income generating capacity and diversification, ownership of agricultural and nonagricultural assets, access to social safety nets and basic services, as well as household stability and adaptive capacity to shocks. By providing a steady and predictable source of income, particularly one that is unconditional, cash transfer programs are hypothesized to impact positively on the productive capacity of households and asset ownership without negatively affecting pre-existing social safety nets and access to basic services. The net effect of this should be improved food security and ability of households to respond to shocks and stressor with more positive coping strategies that are not detrimental to long term development prospects. In this paper, we examine the impacts of the Malawi Social Cash Transfer Progam (SCTP) on the dimensions of resilience and overall resilience score. Our measure of resilience follows the methodology adopted by the Food and Agricultural Organization (FAO) called the RIMA model. We further examine the reliability and validity of the resilience index in predicting future food security. We find significant positive impacts of the SCTP on agricultural and non-agricultural asset ownership, crop production, livestock ownership and household debt situation. We find no crowding out effects of the SCTP on access to private and public social safety nets, and no signs of reduced labour hours although there is some reduction in the hours spent on casual labour. We also find significant positive impacts of about MWK 13,000 on overall per capita consumption as well as a MWK 7900 on per capita food consumption. In addition, we find significant positive effect on household food security, meal frequency, meal quality and dietary diversity. Our estimate of household resilience, using the FAO RIMA II model, also shows significant improvement in the household resilience index for the treatment households. Using the information on actual household coping responses to shocks over the last 12 months, we assess the reliability of the resilience score be examining its predictive power on the coping strategies adopted by households in response to shocks. We find a strong positive association between the resilience index and the share of positive coping responses to shocks. While 37 per cent of households in the lowest quintile of the resilience score are able to adopt positive coping strategies to shocks, the corresponding figure for households in the highest quintile is 71 per cent, with noticeable difference between T and C households. We also examine the predictive power of the resilience score to food security using only the sample of C households in order to exclude the effect of the SCTP. The results of this analysis also show that high resilience score at baseline was reasonably predictive of food security at endline among the C households, indicating a reasonable level of reliability of the resilience score. While the SCTP had no explicit objective on resilience, the overwhelming evidence of increased resilience and the association thereof with actual positive coping strategies to shocks experienced by the households suggests that households that benefit from unconditional cash transfer programs are able to make the right decisions that contribute to building household resilience in the many dimensions it is construed. Keywords: Resilience, RIMA II, Cash Transfer, Social Protection, Food Security.

3 1. Background The concept of Resilience is becoming increasingly popular within the international development community as a framework for profiling and ranking households in terms of their response capacity to shocks and stressors to livelihoods, particularly those that threaten food security. The objective is to provide a single reference indicator for summarizing multidimensional aspects of household livelihoods in order to better inform development and humanitarian interventions and also summarize program impacts. The term has a long history of use in mental health studies where resilience is defined as the ability to withstand and rebound from disruptive life challenges. In the development literature, resilience is discussed in relation to threats to livelihoods, often occasioned by shocks that can be natural or man-made, exogenous or endogenous, seasonal or recurrent, short or protracted (D Errico et al. 2013; FAO II 2014). The definition of resilience in the development literature is still a matter of some discussion due to the multidimensional nature of the term, and contemporary definitions differ mainly in terms of scope and emphasis on the types of threats to livelihoods that have to be taken into consideration. The Resilience Alliance defines the resilience as The capacity of a system to absorb disturbance and reorganise while undergoing change. DFID defines it as the ability of countries, communities and households to manage change, by maintaining or transforming living standards in the face of shocks or stresses such as earthquakes, drought or violent conflict without compromising their long-term prospects, while the FAO s Resilience Measurement Technical Working Group defines it as the capacity that ensures adverse stressors and shocks do not have long-lasting adverse development consequences. (Resilience Alliance, 2002). Barrett and Constas (2014) define development resilience as the capacity over time of a person, household or other aggregate unit to avoid poverty in the face of various stressors and in the wake of myriad shocks. If and only if that capacity is and remains high over time, then the unit is resilient. The common thread through these and other definitions is the notion that resiliency reflects an ability to successfully avoid poverty and food insecurity even in the event of negative shocks or stressors to an established pattern of livelihood. The relevance of this concept cannot be overemphasized due to the increasing disruption in food supplies and agricultural productivity caused by climate change, as well as the frequent outbreaks of civil unrest and armed conflict. Conceptually, a more resilient household is one that is better able to anticipate and manage its exposure to negative shocks to livelihood, and when preventive measures fail, be able to withstand with more positive coping strategies. For example, households that make use of irrigation or other soil management techniques in farming are generally better positioned to avert the full effect of droughts, and also more likely to have higher productivity that minimizes the risk of food insecurity. Efforts to measure resilience are still very much debated both theoretically and empirically. However, there seems to be general consensus that a household s resilience encompasses aspects of household income generating capacity and diversification, ownership of agricultural and non-agricultural assets, access to social safety nets and basic services, as well as household stability and adaptive capacity to shocks. By providing a steady and predictable source of income, particularly one that is unconditional, the SCTP is hypothesised to positively impact on household income generation capacity, ownership of assets and household human capital such as health and education. We also hypothesise that the SCTP would not negatively impact on pre-existing access to social safety nets and basic services, or household demographic composition. The net effect of these effects should result in improved food security, lower exposure to the effects of perennial or seasonal shocks, and strengthened households ability to cope with negative shocks with more positive coping strategies that do not undermine long term development objectives. This expected outcome is not automatic or guaranteed. The use to which households put the SCTP money determines how much they can improve on their livelihood and ability to manage shocks and stressors to livelihoods. The choice of investments can also be constrained by the household s preexisting conditions as households with tighter food consumption budget constraints may not be able to make medium to long term productive investments or asset accumulation to improve their resilience. This report examines the impacts of the SCTP on household resilience and provide some validity test of the resilience score by analysing the relationship between the resilience score and the use of 1

4 positive coping strategies in response to shocks. We also examine the predictive power of the resilience score for use as a ranking tool by examining the relationship between endline food security and baseline resilience for the control households who had no exposure to the SCTP treatment. The next section of provides an overview of the SCTP programme followed by a description of the broader impact evaluation design and the data source for the analysis. Section four provides the broad intent-to-treat (ITT) impact estimates on the various dimensions of resilience. Section five provides a description and estimation of household resilience capacity index using the FAO RIMA II model, and analysis the program impacts and the validity tests described above. Section six provides a summary and conclusion. 2. Overview of the Malawi SCTP Programme The Malawi Social Cash Transfer Programme (SCTP) is one of the several cash transfer programs currently being implemented by governments and development partners across Africa. Locally known as the Mtukula Pakhomo, the SCTP is an unconditional cash transfer programme targeted to ultrapoor, labour-constrained households. The programme began as a pilot in Mchinji district in 2006 and is run by the Government of Malawi (GoM). Since 2009, the programme has expanded to reach 18 out of 28 districts in Malawi. The programme has experienced impressive growth beginning in 2012, and most notably in the last two years. By December 2015, the SCTP had reached over 163,000 beneficiary households. The objectives of the SCTP are to reduce poverty and hunger, and to increase school enrolment rates in these ultra-poor households. The first evaluation of the programme, the impact evaluation of the pilot project in Mchinji, demonstrated that the Malawi SCT Pilot Scheme had a range of positive outcomes including increased food security, ownership of agricultural tools and curative care seeking. 1 Since that time, the programme has witnessed some changes in targeting and operations, and significant expansion. The expectation is that these improvements will lead to even stronger impacts for the larger target population. The SCTP is administered by the Ministry of Gender, Children, Disability and Social Welfare (MoGCDSW) with additional policy oversight provided by the Ministry of Finance, Economic Planning and Development (MoFEPD). UNICEF Malawi provides technical support and guidance. Funding for the programme from was largely provided by the Global Fund to Fight AIDS, Tuberculosis and Malaria (GF). In 2011, the German Government (through Kreditanstalt für Wiederaufbau, or KfW) and the GoM signed an agreement to provide substantial funding for paying arrears in existing areas. In 2013, Irish Aid signed an agreement to expand into one new district, and in 2014, KfW and the European Union (EU) topped-up donor contributions to enable full coverage in the seven existing districts, as well as scale-up into eight additional districts. Also in 2014, GoM launched a government-funded district (Thyolo) and the World Bank committed to providing resources to expand into two additional districts. The SCTP was launched in these 11 newly funded districts starting in mid-2014 through early 2015, bringing coverage to 18 districts. Eligibility criteria are based on a household being ultra-poor (unable to meet the most basic urgent needs, including food and essential non-food items such as soap and clothing) and labourconstrained (defined as having no member fit to work or having the ratio of not fit to work to fit to work of more than three). Household members are defined as unfit to work if they are below 19 or above 64 years of age, or if they are aged 19 to 64 but have a chronic illness or disability, or are otherwise unable to work. 2 Beneficiary selection is done through a community-based approach with oversight provided by the local District Commissioner s (DC s) Office and the District Social Welfare Office (DSWO). Community members are appointed to the Community Social Support Committee (CSSC), and the CSSC is responsible for identifying households that meet these criteria and creating a list. These lists are to include roughly 12 per cent of the households in each Village Cluster (VC), and 1 Miller, C., Tsoka, M., & Reichert, K. (2010). Impacts on children of cash transfers in Malawi. In S. Handa, S. Devereux, & D. Webb, Social protection for Africa's children. London: Routledge Press. 2 Social Cash Transfer Inception Report, Ayala Consulting. July

5 after further screening, the list is narrowed in order to achieve a target coverage rate of 10 per cent. The ultra-poor eligibility condition is implemented through a proxy means test (PMT). The transfer amount varies based on household size and there is a schooling bonus determined by the number of children in the household enrolled in primary and secondary school. Transfer amounts were updated just prior to the start of this evaluation in Due to inflation and decline of the value of the real transfer, transfer amounts were increased again in May The transfer amounts are shown in Table Table 2.1.1: Structure and Level of Transfers (Current MWK) 2013 to May 2015 After May Member 1,000 1,700 2 Members 1,500 2,200 3 Members 1,950 2, Members 2,400 3,700 Each primary school child Each secondary school member ,000 1 Provided for household residents age 21 or below in primary school. 2 Provided for household residents age 30 or below in secondary. To put these amounts in perspective, Table shows the average transfer payment and transfer as share of the household baseline consumption. On average, the total annual transfer amount received by households was MWK25,622 and the average monthly per capita of the transfer was MWK 559. We find that on average, the transfer represented 20 per cent of baseline consumption among all beneficiaries, but was higher at 27 per cent among the poorest 50 per cent of households at baseline. Additional details of the implementation and operational performance can be found in the main impact evaluation report (Handa et al, 2016). In particular, there was high adherence in terms of disbursement with up to 99 per cent of target beneficiaries receiving payments as expected. The quantum of money received was also generally consistent with the schedule in Table except for lack of adjustment for rolling household size. There was little reference to corruption in terms of program officers demanding payments from recipients, and recipients were generally satisfied with the mode of payment. Although there were some misconceptions about eligibility for receiving the SCTP, perceived conditionalities regarding the expenditure of the SCTP money, how long into the future beneficiaries expect to receive the transfer, and delays encountered in going to receive the transfer, there is reason to believe that treatment has been very successful for which reason we would expect to see the theorized impacts. Table 2.1.2: Average Transfer Payment and Transfer Share Total Midline Poorest 50 per cent Small hhld Large hhld Female head Household Size Real hhld total annual transfer (MWK) 22,310 24,300 19,016 25,855 22,486 Real PC total monthly transfer (MWK) Real transfer share Proportion of hhlds with transfer share < 20 per cent N 1, ,361 3

6 Endline Household Size Real hhld total annual transfer (MWK) 25,622 28,180 21,347 29,663 25,697 Real PC total monthly transfer (MWK) Real transfer share Proportion of hhlds with transfer share < 20 per cent N 1, Notes: Transfer values expressed in real August 2013 national prices, MWK. Small households contain four or fewer members. Descriptive statistics are corrected for multi-stage survey design. 3. Impact Evaluation Design This section provides key highlights of the impact evaluation design and the analytical framework. Additional details can be found in main impact evaluation report (Handa et al, 2016). 3.1 Study Design The impact evaluation for Malawi s SCTP uses a mixed method, longitudinal, experimental study design, combining quantitative surveys, qualitative interviews and group discussions, and simulation models to demonstrate wider community economic impacts. 3 The study districts, Salima and Mangochi, were selected for the study in order to integrate with GoM s SCTP expansion plans. The MoGCDSW had plans to conduct retargeting in existing programme areas, and to expand the SCTP to cover 18 districts, starting in The districts scheduled for scale-up in early 2013 were Salima and Mangochi, so the MoGCDSW took this opportunity to integrate an impact evaluation into the planned expansion activities. Subsequently, the research team worked with MoGCDSW, Ayala Consulting and development partners to randomly select two study Traditional Authorities (TAs) in each district (Maganga and Ndindi TAs in Salima, and Jalasi and M bwana Nyambi TAs in Mangochi). The quantitative survey design consists of a cluster-randomized longitudinal study with baseline surveys (household, community and business) which began in July 2013 and two follow-up surveys (household and community) the midline survey was conducted starting in November 2014 and the endline survey was conducted starting in October The qualitative survey is an embedded longitudinal study of 16 treatment households, which includes three main components: in-depth interviews (IDIs) with the caregiver and a young person (aged at baseline) from each household at baseline and follow-up; key informant interviews (KIIs) with community members at follow-up; and focus group discussions (FGDs) in each study TA at baseline and follow-up. Insights from these qualitative interviews and discussions with community members provide complementary data to that obtained through the surveys and will allow us to examine certain topics in more depth, in particular, the role and evolution of social networks and the mechanisms and dynamics that shape outcomes related to the cash transfer programme. Baseline data collection was conducted to allow the study team to accurately describe characteristics of beneficiary households before receiving any cash transfers. Midline and endline data has been compared to data collected at baseline using a difference-in-differences (DD) estimation approach to assess the full impacts of the SCTP. Data collected on the control group allows the researchers to identify which impacts over time are directly attributable to the cash transfer, controlling for outside influences. This is done by taking the overall changes experienced by 3 The FAO, with direct funding from the Department for International Development-United Kingdom (DFID-UK), built a simulation model to predict the potential of the SCTP to generate local economy-wide effects. Those results are reported separately in: Thome, K., Taylor, J.E., Tsoka, M., Mvula, P., Davis, B. and Handa, S., Local Economy-wide Impact Evaluation (LEWIE) of Malawi's Social Cash Transfer (SCT) Programme, PtoP project report, FAO - March

7 beneficiaries and subtracting the changes also experienced by control households. The difference in these two are attributed to the programme and considered programme impacts. 3.2 Sampling and Data Collection The sample for the quantitative longitudinal impact evaluation includes 3,531 SCTP-eligible households and 821 non-eligibles located in 29 VCs across the four TAs in the two districts at baseline. There are 14 VCs (1,678 households) in the treatment (T) group and 15 VCs (1,853 households) in the control (C) or delayed-entry group. Data on the non-eligible households were collected to enable FAO to build the local economy simulation model. 6 The study design uses both random selection (for the selection of study areas at the TA and VC level) and random assignment (to determine T and C VCs), the most rigorous approach available according to evaluation literature. 4 This randomization was done in cooperation with GoM, and was a transparent process open to the public, and the assignment to T-C status was public and attended by local community leaders. The baseline data was used to check for balance between T and C households in order to assess the performance of the randomization and the results showed that T and C households were balanced on more than 100 relevant variables that were examined. After treatment and control VCs were assigned, the qualitative sample of 16 households was selected from treatment VCs for IDIs of the caregiver and a young person. We used a stratified sampling approach to facilitate comparison across sex and orphan status, resulting in a sample that was half male and half orphaned. Geographically, our sample covers two districts, Salima and Mangochi, and four TAs (Salima Maganga and Ndindi TAs; Mangochi Jalasi and M bwana Nyambi TAs). Four households were selected from each TA. We determined the sample size based on our previous experience, guidelines for longitudinal qualitative research, and feasibility. A prerequisite for selection of a household was that the household had to have at least one youth aged years of age (at the time of baseline) who had completed the Young Person s Module in the quantitative survey. This allows for a richer analysis of the youth IDIs, as the qualitative interview could be linked to information on behaviour and attitudes of this same youth from the quantitative survey. These households were then sorted based on gender and age of caregiver and young person, and other characteristics of the young person. Sixteen households were selected on the basis of having a balance of characteristics among the youth respondents, including female/ male, orphan/ nonorphan, had sex/ never had sex and currently enrolled in school/ not currently enrolled in school. Alternate households with similar characteristics were selected to match each of the 16 selected, in case participants refused the IDI or were unavailable. Focus group discussions (FGDs) at midline were held with two separate groups (beneficiaries and non-beneficiaries) in each of the four TAs, for a total of 10 FGDs. 5 The groups were divided into programme beneficiaries and community members not receiving the transfer in order to allow participants to speak freely, without stigma or judgement from the other group. FGD participants were community members aged 18 and above who have detailed knowledge of the community and were invited by the local village heads. The number of FGDs was determined by the fact that we wanted to cover each TA to account for general geographical and cultural differences that could affect the impacts, perceptions, and operations of the SCTP. The specific locations within the TAs were driven by the fact that, for logistical purposes, the FGDs were conducted during the same time period as the IDIs; therefore, FGDs were held in the same VCs where the IDIs were given. The survey instruments used consists of six major components: 1. Household Survey administered to the main respondent for the household; 2. Young Person s Module for up to three youth ages in the household (age at endline); 3. Anthropometric Measures for children ages 6 months to 71 months in the study households; 4. Community Survey given to a group of knowledgeable community members to gather information on community norms, resources, pricing and access to services; 4 Shadish WR, Cook TD, Campbell DT. Experimental and Quasi-Experimental Designs for Generalized Causal Inference. Boston: Houghton-Mifflin An additional set of FGDs was conducted in Mangochi since time permitted the team to do so. 5

8 5. IDIs for caregiver and one youth from 16 treatment households; 6. KIIs and FGDs with knowledgeable community members to discuss impacts, perceptions, and operations of the SCTP. Beneficiary and non-beneficiary FGDs were held separately. Survey instruments were reviewed for ethical considerations and approved by the UNC Internal Review Board (IRB) and Malawi s National Commission for Science and Technology (NCST), National Committee for Research in Social Sciences and Humanities (UNC IRB Study No ; Malawi NCST Study No. RTT/2/20). Instruments are available online at: Attrition Attrition occurs when households from the baseline sample are missing in the follow-up surveys. There are different reasons for households not responding in subsequent survey waves. Migration, death, separation, or the dissolution of households can cause attrition and make it difficult to locate a household in the second or third wave of data collection. Attrition can cause problems for an evaluation because it not only decreases the sample size (leading to less precise estimates of programme impact), but it could also introduce bias into the analytic sample. If attrition is selective, it could lead to incorrect programme impact estimates, or it could change the characteristics of the sample and therefore, it could affect the representativeness of the impact results. There are two types of attrition: differential and overall. Differential attrition occurs when the treatment and control samples differ in the types of households or individuals who leave the sample. Differential attrition can create biased samples by reducing or eliminating the balance between the T and C groups achieved at baseline. Since we will conduct the analysis using the households present in all three waves of the survey, it is also important to examine for overall attrition, which is the total share of observations missing at the follow-up surveys from the original baseline sample. Overall attrition can change the characteristics of the remaining sample of analysis and render it nonrepresentative of the population from which it was obtained. Overall attrition can affect the ability of the study s findings to be generalized to the population of interest. Ideally, both types of attrition should be null or small. We investigated attrition at endline for the quantitative sample by testing for similarities at baseline between (1) treatment and control groups for all households included in the panel of households, that is, for the households interviewed at baseline and in both follow-up surveys (differential attrition) and, (2) all households in the panel and the households who were missing in either the midline or the endline survey (overall attrition). Fortunately, we do not find evidence of differential attrition, meaning that we preserve the balance between the T and C groups found in the baseline survey. Summary attrition tables are given in Appendix A. However, there is evidence of overall attrition in the sample which we correct for by using modelled inverse probability weights. The attrition rates and effective sample sizes are shown in Table Table 3.3.1: Household In the Panel and Attrition Rates by T - C Status and District In Panel Rate (Per cent) Attrition Rate (Per cent) N Total sample ,531 Treatment group ,678 group ,853 District Status Salima Treatment Salima Mangochi Treatment Mangochi

9 For the qualitative sample, the caregiver and one youth, aged from 16 households were interviewed at baseline, for a total of 32 participants. At midline, three female youth had left their homes for marriage, and one went to live with relatives. One male youth left home to attend secondary school in another district. While these five youth were no longer in the SCTP households at follow-up, the research team was able to trace all of them for the follow-up interviews. One caregiver, a grandmother, had passed away shortly before midline interviews and the youth had gone to live at his aunt s house. Both the youth and the aunt were interviewed at midline. Therefore, at midline, 32 interviews were conducted, and 31 of those were with the same baseline participants, the only exception being the deceased participant. Our team had similar success with retention at endline; while six youth (three boys, three girls) were no longer living at the households where they were initially recruited, the interviewers were able to track and interview all of them. Of note, among the six who had left their households, all three females had married while all three males had left to study (two in secondary, one in madrasa). Three females who had married had returned home by endline and were interviewed in their original households. Overall, 32 interviews were conducted at endline with the same 32 respondents from midline. 4. Program Impact on Resilience Domains This section presents the program impacts on the various domains of resilience. The domains include economic activities, asset ownership, access to credits and transfers, access to social safety nets, labour use, shocks and coping, consumption and food security. Impacts are estimated using DD regression and are reported as average treatment effects. 4.1 Impact on Economic Activities One of the objectives of the SCTP is to reduce poverty and hunger among beneficiaries. Since household poverty and hunger are invariably the result of household production being in deficit of household demand, we recognize that increasing household production is the more sustainable way to reduce poverty and hunger in the long term. The SCTP cash is hypothesized to act as a catalyst for behavioural responses and necessary investments in household economic activities that will result in increased production. Our analysis shows an impact of 62kg in overall crop harvest, driven mainly by an impact of 60kg on the five main staple crops (maize, groundnut, rice, pigeon pea and pumpkin). There is also an impact of MWK 12,000 on the total value of crop harvest. On livestock production, the proportion of T households involved in livestock production at baseline more than doubled at endline (from 29 per cent to 59 per cent) and the impact on raising livestock was 22 pp. Livestock owned, measured in terms of the standard tropical livestock unit (TLU) equivalents also more than doubled among T households from baseline to endline, and the impact on this indicator was about 5 pp. We also found significant positive impacts on livestock consumption, and expenditure on livestock purchases over the past 12 months. We generally do not find any impacts on the operation of non-farm household enterprises (NFE) or on enterprise profitability. Overall, we find an impact of 0.24 units in the number of economic activities that households are engaged in, an indicator of income source diversification and strengthening (Table 4.1.1). 7

10 Table 4.1.1: Summary Impacts on Economic Activities Variable Impact Impact Diff (EL-ML) Crop production household Total crop harvest (kg) Total crop harvest (kg) - Staples Total value of crop harvest (MWK) Raised or owned livestock TLU owned presently Household has non-farm enterprise Number of economic activities ** (1.11) (-0.86) (2.41) *** *** (5.07) (0.85) (3.18) *** *** (4.73) (0.71) (3.23) 12, *** , *** 29, , , (3.80) (0.12) (3.86) 0.220*** 0.135*** 0.084*** (5.15) (2.78) (3.24) 0.051*** 0.034** (3.73) (2.47) (1.67) * (0.28) (-1.36) (1.78) 0.246*** *** (3.98) (1.45) (3.53) N 9,902 9,902 1,576 1,575 1,726 Notes: Estimations use difference-in-differences modelling among panel households. Binary outcomes are estimated using LPM. All estimations control for baseline head of household s characteristics (age in years, sex, indicator of any schooling, indicator of literacy, marital status), household demographic composition and size, indicators for new household members and household member outmigration, and a vector of contemporaneous cluster level prices. Robust t statistics were obtained clustering at the different levels of the sampling design and are shown in parenthesis. * 10% significance ** 5% The main impact evaluation report (Handa et. al, 2016) has extensive coverage on various aspects of these household economic activities including impacts inputs into crop production (fertilizer use, farm size, etc), crop sales, livestock consumption and sales, and the specific livestock types (goat/sheep, chicken, duck/geese). We also present some of the heterogeneous treatment effects on these indicators. In particular, we find similar effects, often with larger coefficient sizes, for the baseline bottom 50 per cent of households. Annex B of this report provides some of the activity specific and heterogeneous impact tables. 4.2 Impacts on Asset Ownership We investigate the impacts of the SCTP on ownership and investments in agricultural and nonagricultural assets. At baseline, about 93 of households owned or cultivated land, and the inability to own basic farming tools often led to borrowing or renting of assets, taking away from already scarce household resources and reducing productivity. Ownership of basic durable goods is indicative of improved quality of life and also serves as a store of wealth that can be sold or pawned to deal with 8

11 emergencies arising out of shocks or stressors to livelihood. Tables and provide a summary of the impacts on ownership of assets. Table is based on cross-sectional differences for midline and endline since the information on asset ownership was not collected at baseline. In either case, we find significant positive impacts on a household wealth index based on the first principal component for the ownership of the agricultural or non-agricultural assets. We also find significant impacts on asset purchases in the last twelve months as well as the monetary value of purchases. Details on the specific assets purchased and of the heterogeneous impacts are provided in main evaluation report by Handa et al, Table 4.2.1: Impacts on Ownership and Purchases of Agricultural Assets Diff Variable Impact Impact (EL-ML) Own any asset 0.065*** *** (3.21) (0.60) (2.79) Number of asset types 0.249** * (2.41) (0.68) (1.70) Asset ownership index 0.302** (2.68) (0.91) (1.64) Any Asset Purchase in last 12m 0.072** 0.089*** (2.76) (3.40) (0.72) Total expenditure on purchases (MWK) * ** (2.02) (2.11) (0.36) N 9,901 9,901 1,576 1,574 1,726 9

12 Table 4.2.2: Impacts on Ownership and Purchases of Durable Goods Dependent Endline Midline Midline Midline Endline Endline Variable Impact Impact Treatment (1) (2) (3) (4) (5) (6) Owns any durable good 0.141*** 0.085*** (7.50) (3.74) Number of durable 0.619*** 0.124* goods owned (5.21) (1.87) Durable good 0.326*** ownership index (3.47) (1.20) Any expenditure on 0.068*** 0.029*** goods in last 12 months (3.07) (4.55) Expenditure on durable ** *** goods in last 12 months (MWK) (2.17) (3.01) N 3,300 3,299 1,574 1,725 1,574 1,726 Notes: Coefficients represent cross-sectional differences between panel T and C households at Midline and at Endline. Binary outcomes are estimated using LPM. See Table for additional explanatory notes on model specification, including a list of control variables utilized. * 10% significance ** 5% 4.3 Impacts on Access to Credit and Transfers Access to credit and other transfers is another important dimension to household livelihood. Credits and transfers could be relied upon to smoothen consumption and other expenditure in times deficit. This could be during the lean agricultural season or illness of household members. Credits and transfers could also be necessary for occasional large expenses such as payment of school fees at the start of the school year, or investment in equipment for a non-farm business. Borrowing and purchases on credit could prove regressive especially if they come with high-interest payments and are used directly for consumption. By providing unconditional regular cash to the households, the SCTP is expected to ease the demand for credits, especially for consumption. At the same time, it is possible that being enrolled in the SCTP could extricate beneficiaries from networks of friends and relatives who would otherwise provide credit or other types of support. Additionally, beneficiaries may often be obligated by social norms to share their money with other friends and relatives through increased out-transfers. The net effect of all these dynamics can have profound effects on how the SCTP improves the livelihood of beneficiaries. The survey instrument therefore elicited information on various aspects of credit and transfer activities and behaviour in all three waves. Questions were asked about outstanding debts that originated more than 12 months prior to each survey round, as well as loans and credit purchases in the 12 month period preceding each data collection. Our analysis shows a five pp impact reduction in the proportion of households with a debt on a loan that originated more than 12 months prior to the survey. We also find a nine pp impact reduction in purchases on credit and a further seven pp impact reduction on the proportion of credit purchases that have been fully repaid. We find no impacts on the taking a loan in the last 12 months or fully repaying the loan taken (Table 4.3.1). Putting it all together, a household was in debt if it had outstanding balances from more than 12 months ago, or had not fully repaid any loan or credit purchases (including any accruing interest) taken in the past 12 months. Overall, we find a 10 pp impact reduction on the proportion of households in debt, and a comparative decrease of MWK 916 in the total debt in T households. 10

13 Table 4.3.1: Impacts on Loans and Credits Variable Impact Impact Diff (EL-ML) Still owes on loan from 12+ months Took a loan in last 12m ** ** (-2.69) (-0.80) (2.24) (-0.34) (-1.22) (0.76) Loan fully paid (0.83) (1.44) (0.36) Purchased on credit in last 12m Credit on purchases fully paid ** ** (-2.34) (-2.52) (0.54) 0.072*** 0.049** (2.88) (2.52) (1.03) Currently Owes ** ** (-2.66) (-2.59) (0.74) Total current debt (MWK) *** ** ** , , (-3.22) (-2.43) (2.18) N 9,902 9,902 1,576 1,575 1,726 Further to the positive outlook on household debts, we investigate if this is caused by differential credit constraints. The results in Table could be observed if T households were more likely to be refused loans or credits when in fact they needed it and actually applied for it. Additionally, if T households did not seek a loan or seek to purchase on credit because they were sure they would be refused, then we could still get the results in Table There would be some concern if either of these reasons contributes significantly to the results in Table There were question in the survey instrument to interrogate all these mechanisms, and our estimations show that T households were significantly less likely to have been refused a loan they applied for, or denied to buy on credit. We also find null effects on the baseline situation regarding access to sources of credit purchase and loans. Our overall indicator on credits is household credit constraint. A household is considered credit constrained if the household: a. has a loan debt, but actually wanted more loan than it received at the same interest rate; or b. would ask for a loan or purchase on credit if they were sure they could get it; or c. has been refused a loan or denied a purchase on credit when they actually asked. This does not control for whether they actually needed a loan or credit, but rather whether they had any barriers in case they needed it. We find no significant impact on this overall indicator (Table 4.3.2). In reconciling this with the result in Table 4.3.1, we can be quite sure that the positive outlook on credit among T households is not likely a result of differential credit constraints, but more likely a result of lack of need for credit. We also recognize that a credit is not necessarily a bad thing, especially if it can be put to productive use to generate multiplying effects. We accordingly examine impacts on the purpose for obtaining a loan or credit and find an eight pp impact decline in the share of household using credit for consumption. The impacts on use of credit or loan for health, education and productive investments are all null (Table 4.3.3). 11

14 Table 4.3.2: Impacts on Credit Constraints Variable Impact Impact Diff (EL-ML) Applied for loan but refused ** (-2.17) (-1.00) (0.48) Asked to buy on credit but refused ** (-2.19) (-0.71) (1.53) Wanted larger loan at same interest rate (-0.45) (-0.85) (0.36) Sure to get a loan if applied (-1.27) (-1.01) (0.63) Would apply for loan if sure can get * (-1.49) (-0.46) (1.76) Sure can buy on credit if asked (-0.20) (0.91) (0.74) Would ask to purchase on credit if sure can get (0.40) (0.35) (0.10) Loan/Credit Purchase constrained (0.96) (0.50) (0.50) N 9,902 9,902 1,576 1,575 1,726 Table 4.3.3: Impacts on Credit Use Variable Impact Impact Diff (EL-ML) Some loan used for prod. invest (0.33) (-0.49) (0.75) Some loan used for consumption (-0.30) (-1.08) (0.46) Some loan used for education (-0.88) (-0.12) (0.96) Some loan used for health Some credit used for prod. invest Some credit used for consumption (-0.84) (-0.39) (0.67) (1.07) (0.81) (0.18) ** *** (-2.27) (-2.84) (0.13) Some credit used for education (-0.84) (0.07) (0.80) Some credit used for health (-1.23) (-0.06) (1.36) N 9,902 9,902 1,576 1,575 1,726 12

15 On transfers in and out of the household, we examine transfers of cash, food or labour. We find no impacts on any in- or out transfers, both at the intensive and extensive margins (Table 4.3.4). We also investigate the question of whether households could get any such support when they actually needed it, and also found no impacts (Table 4.3.5). It can thus be argued that the SCTP does not induce a crowding out of pre-existing sources of in-transfers or excess demand for out-transfers. Table 4.3.4: Impacts on In- and Out-Transfers Variable Impact Impact Diff (EL- ML) (1) (2) (3)=(1)- (2) (4) (5) (6) Any in-transfer of cash, food or labour Any out-transfer of cash, food or labour Total value of cash, food of labour in-transfer (MWK) Total value of cash, food of labour out-transfer (MWK) Net transfer of cash, food or labour (MWK) (1.16) (0.03) (1.17) (1.44) (-0.10) (1.52) , , , , (0.55) (0.69) (0.11) ** , (1.41) (-0.36) (2.12) , , , , (0.24) (0.84) (0.64) N 9,899 9,899 1,576 1,575 1,726 Table 4.3.5: Perceived Availability of Support Variable Impact Impact Diff (EL-ML) Household sure can get Cash Transfer in case of need Household sure can get Food/Other Consumables in case of need Household sure can get Labour or Time in case of need Household sure can get Agric Implements/Inputs in case of need (1.03) (0.70) (0.33) (0.76) (-0.07) (0.86) (0.07) (-0.62) (0.86) (0.35) (-0.34) (0.77) N 9,898 9,898 1,576 1,575 1,726 13

16 4.4 Impacts on Access to Social Safety Nets Apart from individuals, the GoM and other non-governmental organizations also provide various social safety nets (SSN) to which poor households have access. It is also desirable that the SCTP will not have any crowding-out effect on the access to these social safety nets. To derive the most benefit from the SCTP, it is essential that the cash transfers act as a complement to these networks and social safety nets, not as a substitute. Table shows the impacts of the SCTP on access to social safety nets. Overall, we do not find any impacts on benefiting from at least one SSN or on the number of SSNs households benefit from. We also do not find an impact on the value of the SSN benefits received, nor on benefits from the voucher for fertilizer program (FISP) a flagship government program to boost agricultural productivity through fertilizer use. Table 4.4.1: Impacts on Social Safety Nets Variable Impact Impact Diff (EL-ML) Any SSN benefit (-0.74) (-0.80) (0.11) No. of SSN benefits (-1.53) (-1.31) (0.70) Value of SSN benefits (MWK) , , , Voucher for fertilizer (FISP) Value of Voucher for fertilizer (-0.12) (-0.21) (0.10) (0.42) (-0.13) (0.83) , , , (0.83) (0.36) (0.45) N 9,901 9,901 1,576 1,575 1,726 We analyse the impacts on specific SSNs and find generally null impacts except on the proportion of households that benefit from the other free food program, which has seen a negative 14 pp impact. However, we do not find an impact on the value of free food received which is quite surprising given the huge impact on the extensive margin (Table 4.4.2). 14

17 Table 4.4.2: Impacts on Specific Social Safety Nets Variable Impact Impact Diff (EL-ML) Free maize (-1.26) (-1.15) (0.71) Quantity of Free Maize (kg) (-1.02) (-0.97) (0.25) Other free food ** * (-2.09) (-1.14) (1.99) Value of Other free food (-0.34) (-0.05) (1.19) Food/Cash for work Value of Food/Cash for work (0.46) (-0.87) (1.33) ** (0.05) (-2.10) (1.64) School Feeding (-1.24) (-1.24) (0.45) Value of School Feeding , Community Based Childcare Value of Community Based Childcare 1, (-1.04) (-1.35) (0.23) (0.31) (0.29) (0.12) (0.09) (-0.58) (0.91) N 9,901 9,901 1,576 1,575 1, Impacts on Labour Use The extent to which a household has available labour is likely to play a mediating role on how the SCTP impacts household economic activities and productivity. If labour is available and underutilized due to liquidity or knowledge constraints, an increase in work participation would be expected for less labour-constrained households. This would increase household productivity and create a multiplying effect beyond the size of the SCTP amount. Conversely, households with tighter labour constraints may be less responsive in their work participation if members are not fit to work, and the SCTP cash would go directly into consumption. The more desirable outcome is that households are able to re-allocate labour from less productive activities to more productive ones, and to be able to move away from hazardous labour, particularly for children. Appropriate modules in the surveys allow for analysis of these effects. We first analyze the household labour constraint situation at baseline. A household is defined as severely labour constrained if it has no member fit to work (FTW). A person is considered fit to work if person is aged between 19 and 64 years, and has no chronic illness or disability, or is otherwise unable to work. If a household has at least one member FTW and the ratio of not fit to work (NFTW) members to FTW member is greater than or equal to 3, then the household is considered moderately labour constrained. A household is labour unconstrained only if the ratio of NFTW to FTW members is less three. The labour constraint classification is purely a function of the household s own demography, and it is important to add that this classification does not take into consideration the ability of the household to engage hired labour or rely on exchange labour. Additionally, there are less 15

18 Prop. FTW labour intensive income generating activities which household members with chronic conditions or disability, or who are older than 64 years of age could engage in. Accordingly, analysis of actual labor supply extends beyond labour supplied by those who are FTW. Table shows the distribution of households and household members living in each of these household types at baseline. About 29 per cent of households were moderately labour constrained, but these accounted for 39 per cent of individuals. Severely labour constrained households made up 54 per cent of household count and contained 42 per cent of individuals. Overall, there is balance between treatment and control, and this is discussed in greater detail in the main report. Figure shows the proportion of the sample FTW by age. As expected, the share of FTW decreases with age, and the distribution is essentially identical for T and C. Table 4.5.1: Baseline Labour Constraint Status at Household and Individual Levels Household Individuals Status C T Total C T Total Unconstrained Moderately Constrained Severely Constrained Total Age (years) C T Fig : Share of FTW by Age Next, we examine the impact of the SCTP on household labour constraint. There are a number of pathways through which the SCTP could influence how household labour constrain status would evolve. If SCTP households are able to attract new household members FTW, then this would improve the labour constrain status of the household. For example if a 65 year old single member is now able to attract a caregiver to live with because of the improved financial situation, then the labour constrain status changes immediately from severely constrained to unconstrained. Similarly SCTP households may be more able to avert the departure of household members when they are faced with a shock. The result of such effect on household welfare is ambiguous since there could be both negative and positive effects and the outcome depends on which of the effects dominates. Analysis shows that there were no impacts on the number of household members FTW, number of males FTW, share of households severely labour constrained and share of households labour constrained (moderately or severely). Impacts on the number of female members FTW and share of 16

19 households moderately labour constrained were only marginally significant at the 10 per cent level (Table 4.5.2). Thus, it can be argued that changing labour constraint is not a plausible mechanism through which the SCTP could impact other outcomes. Table 4.5.2: Impacts on Labour Supply Variable Impact Impact Diff (EL-ML) Total Members FTW (0.97) (0.42) (0.70) Males FTW * (-0.73) (-1.83) (1.11) Females FTW 0.062* 0.056* (2.02) (1.88) (0.24) Severely Labour Constrained (-1.27) (-1.34) (0.11) Moderately Labour Constrained 0.038* * (2.02) (0.60) (1.83) Labour Constrained (0.16) (-1.26) (1.50) N 9,906 9,906 1,576 1,576 1,726 We further examine the impacts on labour use for household chores and economic activities at the household level. Chores include time spent collecting water, time spent collecting firewood and time spent taking care of children, cooking or cleaning. We find no impacts on time spent on all household chores, own farm activities, fishing and then livestock activities. However, the number of hours in the last seven days spent on casual part time work reduces from 11 to 6 hours among T households with a significant impact of a four hour reduction. We also find a significant impact of three months decrease on the amount of time spent doing casual labour for others (ganyu work) in the last 12 months. We also find an impact of more than one hour increase in the amount of time spent on work outside of the household (excluding ganyu). Finding no impacts on the time spent on livestock production activities is quite surprising in view of the huge impacts on livestock production at both the extensive and intensive margins. This could be an indication of increasing returns to scale, particularly for households which raised livestock at baseline, or through the use of hired labour. To further explore the dynamics of labour use, we examine labour use for each of the main activities to try and see if there are any shifts that still keep the overall time use unchanged despite the significant increases in crop production. We also examine the possible role for the use of hired labour in this dynamic. Table shows the impacts on household and hired labour use for the various farm activities: land preparation and planting, farm management (weeding, fertilizing, etc) and harvest. Here we find no impacts on household reallocation of labour among the activities, but we find significant impacts on the use of hired labour at both the intensive and extensive margins. There is a three pp impact increase in the proportion of households using hired labour. We also find that hired labour is mostly utilized for land preparation and planting. This is reasonable in view of the fact that land preparation and planting is mostly time bound and requires a lot of upfront input to set the stage for the rest of the season. 17

20 Table 4.5.3: Impacts on Labour Use by Activity Variable Impact Impact Diff (EL-ML) All Chores (Hours Yesterday) (0.20) (0.07) (0.19) Own Farm Activities (Days in Past Season) (0.84) (-0.06) (0.96) Fishing (Days in Last 7 Days) Non-Farm Enterprise (Hours in Last 7 Days) Livestock Activities (Hours in Last 7 Days) Casual, Part time activities (Hours in Last 7 Days) Ganyu Work (Months in last 12 Months) Work Outside Household excluding Ganyu (Hours in Last 7 Days) (-1.10) (-1.15) (0.30) ** (0.35) (-0.78) (2.50) (1.10) (0.19) (0.81) *** ** (-3.51) (-2.16) (0.35) ** ** (-2.75) (-2.59) (0.34) 1.003** (2.07) (1.28) (1.07) N 9,906 9,906 1,576 1,576 1,726 Table 4.5.4: Impacts on Household and Hired Farm Labour Variable Impact Impact Diff (EL-ML) HH Lab. for Land Preparation and Planting (Days in Past Season) HH Lab. for Field Management (Days in Past Season) HH Lab for Harvesting (Days in Past Season) (0.21) (0.16) (0.02) (1.22) (-0.17) (1.47) (1.17) (-0.54) (1.41) Any Hired Farm Labour 0.030** (2.08) (1.13) (1.56) Hired Farm Labour (Days in Past Season Hired Lab for Land Preparation and Planting (Days in Past Season) 0.475* (1.94) (0.36) (1.09) 0.266** (2.23) (0.92) (0.96) 18

21 Hired Lab for Field Management (Days in Past Season) Hired Lab for Harvesting (Days in Past Season) (1.18) (-0.03) (1.11) (0.55) (-0.00) (0.59) N 9,901 9,901 1,576 1,574 1,726 Finally, we examine labour allocation across the various activities by broad age-sex groups, namely males FTW, Females FTW, All Men (aged years), Elderly (men and women aged 64 years or older) and children (males or females aged 6-17 years). Overall, the pattern of labour allocation is very similar to the pattern in Table There are no impacts on labour allocation for all household chores, farm activities, fishing and NFE activities. There is an impact on labour allocation to livestock activities by FTW males, but this is significant only at the 10 per cent level. We also do not find impacts on female withdrawal from casual part time activities, or intensification in work outside the household (excluding ganyu). There is a significant negative impact (positive outcome) in children participation in ganyu work. Table 4.5.5: Endline Impacts on Intra-Household Labour Allocation Dependent Variable Members FTW Males FTW Females FTW All Mem Elderly (64+) Children (6-17) All Chores (Hours Yesterday) (1.10) (0.68) (0.20) (0.20) (0.96) (-0.08) Own Farm Activities (Days in Past Season) (1.02) (0.46) (1.05) (0.81) (1.29) (1.10) Fishing (Days in Last 7 Days) Non-Farm Enterprise (Hours in Last 7 Days) Livestock Activities (Hours in Last 7 Days) Casual, Part time activities (Hours in Last 7 Days) Ganyu Work (Months in last 12 Months) Work Outside Household excluding Ganyu (Hours in Last 7 Days) (-1.00) (-0.84) (-0.12) (-0.47) (-1.06) (-1.21) (0.21) (-0.08) (0.29) (0.40) (-0.75) (0.67) * (1.44) (1.91) (1.04) (1.53) (-0.32) (1.11) ** ** ** *** (-2.37) (-2.06) (-1.54) (-2.27) (-0.82) (-2.86) ** * * ** (-2.09) (-1.89) (-1.94) (-2.25) (-0.88) (-1.67) 0.817** 1.771* * 0.471* (2.09) (1.74) (0.99) (2.00) (1.78) (1.01) N 7,055 2,683 4,372 12,042 6,182 21,618 19

22 4.6 Impacts on Shocks and Coping Perhaps more directly related to the issue of resilience is the actual experience of shocks and how the households cope when they experience such shocks. Respondents were asked whether they were negatively affected by a series of shocks and their response to try and maintain their standard of livelihood. These shocks are categorized as covariate shocks (which typically affect the entire community such as droughts, floods/landslides) and idiosyncratic shocks, which are more household specific (such as death of the main income earner in the household, sickness, theft of money, etc.). Coping to these shocks could usually include a mix of strategies some of which are negative (reducing consumption or sending children out to work), positive (relying on own savings/sctp payment, receiving unconditional help from social networks), or ambiguous depending on the extent of the response (e.g. labour intensification could be positive or negative depending on the initial level and thresholds). In Table 4.6.1, we summarize the impacts of the SCTP on the experience of the aggregate shocks and the use of positive and negative coping strategies. We find no impacts of the SCTP on the experience of any negative shock, and on either covariate or idiosyncratic shocks. This is largely expected since the SCTP cannot per se avert the occurrence of many of the shocks listed. However, consistent with expectation, we find a significant 26 pp impact on the share of positive coping strategies and a significant negative impact of 23 pp on the share of negative coping strategies adopted. These two categories are not necessarily substitutes since households typically employ a mix of strategies. At the endline, we also enquired about whether households had experienced any positive shocks such as an inheritance, better pay/job or death of a chronically ill household member (on whom the household had to make a lot of expenses). We find no cross-sectional difference in the experience of positive shocks between T and C households as well. Table 4.6.1: Impacts on Shocks and Coping Variable Impact Impact Diff (EL-ML) Any Negative Shock (-1.41) (0.39) (1.19) No. of Shocks (-0.29) (0.40) (0.57) Any Covariate Shock (-1.32) (0.27) (1.28) Number of covariate shocks (-0.18) (0.22) (0.37) Any Idiosyncractic Shock (0.04) (0.61) (0.77) Number of idiosyncratic shocks (-0.20) (0.40) (0.88) Share of Positive Coping Strategies 0.259*** 0.152** Share of Negative Coping Strategies (3.74) (2.09) (1.14) *** ** (-4.02) (-1.01) (2.36) N 8,722 8,722 1,508 1,383 1,594 20

23 Table gives the impacts on the specific shocks. We find no impacts on the proportion of households that experienced any of the specific shocks in the 12 month period preceding the surveys. Perhaps the one shock the SCTP could have affected is the death of a household income earner through improved health seeking behaviour, but the incidence of this is quite low and also likely to suffer from ceiling effects. The impacts on the specific coping strategies are given in Table We find a significant negative impact of 20 pp on the proportion of households that had to cope by changing eating pattern (relying on less preferred food options, reducing food proportions or number of meals per day). We also find a five pp impact reduction on the use of borrowing as a coping strategy to shocks. The mix of coping strategies, including the role of SCTP is depicted in Fig Table 4.6.2: Impacts on Specific Shocks Variable Impact Impact Diff (EL-ML) Drought/irregular rains (-1.04) (-0.14) (0.93) Unusually high level of crop/livestock pest/disease (0.59) (0.37) (0.17) Unusually high prices of food Serious illness or accident to household member(s) (0.91) (0.44) (0.19) (-0.15) (0.22) (0.56) Death of household income earner(s) (-0.95) (-0.55) (0.72) N 9,902 9,902 1,576 1,575 1,726 Table 4.6.3: Impacts on Coping Strategies Variable Impact Impact Diff (EL-ML) Did nothing (-1.37) (-0.24) (1.35) Own savings (-1.49) (-1.37) (0.05) R'ced external assistance *** ** (-4.35) (-1.07) (2.48) More work *** *** (-3.54) (-3.50) (0.83) Borrowed ** ** (-2.60) (-1.02) (2.12) Household members moved out (-1.38) (-0.76) (0.27) Changed eating pattern *** *** (-3.21) (-1.24) (3.13) N 8,720 8,720 1,508 1,383 1,594 21

24 % 12.25% 13.13% 13.01% 28.93% 25.75% 24.64% 26.37% 11.00% 11.43% 9.16% 11.18% Baseline (C) Endline (C) 0.00% 9.68% 18.08% 22.46% 20.37% 10.03% 19.39% Baseline (T) 47.75% 6.01% 6.19% 9.15% 12.64% 6.01% 12.25% Endline (T) Fig : Strategies for coping with negative shocks (aggregate shares) 4.7 Impacts on Consumption and Food Security The overarching objective of the SCTP is to mitigate the effects of poverty by ensuring food security and maintaining consumption. Adequate consumption and food security are not only essential for survival, but are also instrumental for wellbeing and particularly important for child growth and development. We estimate the impacts on consumption using total annual per capital consumption at the household level. Table shows the impacts on household consumption expenditures. There is a MWK impact on overall per capita consumption and a MWK 7920 impact on food expenditures. Computations use the national poverty and ultra-poverty lines provided by the National Statistics Office (NSO). Details of the poverty lines and inflation factors to account for the timing of the surveys are found in the main impact evaluation report (Handa et al, 2016). A breakdown of food consumption by the major food groups reveals a decrease in the share of expenditure on cereals and an increase in the share of the expenditure on meats and beverages. This shift may be an indication of a shift in preference, but also reflects a quality-for-quantity substitution that augurs well for household nutritional balance. A simple measure of dietary diversity a count of the number of the broad categories a household meal typically comes from shows significant increase in dietary diversity (Table 4.7.2). We also find a significant positive impact on the food consumption score (FCS) a composite score based on dietary diversity and the relative nutritional importance of different food groups. Finally, there is also a significant positive impact on the Simpson s Index of Dietary diversity an index that takes into account not only the count of the food groups, but also the expenditure shares allocated to each group. The computations of the FCS and the Simpson s diversity index follows WFP and FAO methodology 6. 6 See for example: Elliot Vhurumku: Food Security Indicators - Integrating Nutrition and Food Security Programming for Emergency response workshop, 25 to 17 February